TempShap introduces a Shapley-based framework that attributes importance to individual time‑step and feature cells in a time‑series forecast, providing both temporal localization and horizon‑impact profiles. Unlike traditional Shapley explainers that treat each feature as a single player, TempShap’s cell‑level approach captures when and how much a feature matters across the forecast horizon. The method is validated through planted temporal structures, exact cell games, and a closed‑form oracle for linear forecasters, and demonstrates improved deletion‑ranking performance over existing estimators.
By Chi Zhao, Anastasiia Chelnokova, Jing Liu, Elena Parilina
GRAIL is a region‑based framework for explaining time‑series models that moves beyond per‑timestep saliency by treating domain‑specific intervals—such as the P‑wave and QRS complex in ECG data—as players in a cooperative game. It describes each region with human‑readable features, generates counterfactual baselines using an on‑manifold conditional generator, and incorporates static covariates as singleton players to directly weigh tabular evidence against the signal. The method attributes predictions via Owen values with variance reduction, and experiments on synthetic and real data demonstrate faithful, robust explanations that provide meaningful knowledge to domain experts.
By Enver Menadjiev, Yiqiu Shen, Jaesik Choi, Julia Stoyanovich
The paper investigates the phenomenon of explanation multiplicity in SHAP, showing that SHAP explanations can vary significantly across repeated runs even when the model, input, and prediction remain unchanged. The authors introduce a dual‑seed evaluation protocol, a hierarchy of metrics, and null models to quantify this variability, finding that multiplicity is widespread across datasets, models, and sampling strategies. They also demonstrate that common metrics like L2 distance may underestimate instability, while rank‑based metrics reveal substantial changes in top‑ranked features, and that improved sampling methods do not fully eliminate run‑to‑run variation.
By Hyunseung Hwang, Seungeun Lee, Lucas Rosenblatt, Steven Euijong Whang, Julia Stoyanovich
arXiv:2607. 05484v1 Announce Type: cross Abstract: The adoption of non-parametric machine learning models for regulatory capital estimation introduces a fundamental governance challenge: the inability to explain model outputs in a manner auditable by supervisory bodies.
By Ujjwala Vadrevu
arXiv:2512. 03578v3 Announce Type: replace-cross Abstract: Time series extrinsic regression (TSER) refers to the task of predicting a continuous target variable from an input time series.
By Florent Forest, Amaury Wei, Olga Fink
DynSHAP is a SHAP-based framework designed for dynamic survival analysis, extending marginal SHAP estimators to handle time–feature pairs as players in the Shapley game. It introduces Temporal DynSHAP, which models linear dependencies across time and uses conditional sampling to improve explanations. Experiments on synthetic data and two real-world clinical datasets show that Temporal DynSHAP more accurately recovers temporally dependent features and provides faithful attributions for two DSA architectures, enabling medical experts to identify which patient information influenced predictions and when.
By Nastasya Anokhina, Jonas J\"ur{\ss}, Pietro Li\`o
arXiv:2609.24677v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly used to make predictions from numerical time-series histories and textual events. Yet accuracy alone cann...
By Jie Gong, Maowei Jiang, Zhiwei Liu, Yankai Chen, Guojun Xiong, Xue Liu, Min Peng, Qianqian Xie, Sophia Ananiadou
The paper introduces a Hilbert‑valued functional decomposition framework that extends feature‑based explanations to time‑dependent outputs, such as trajectories in demand forecasting. It generalizes functional decomposition to Hilbert‑valued prediction functions and incorporates kernel‑based output representations, allowing explanations at multiple temporal granularities—time‑specific, time‑resolved, and time‑aggregated. The authors validate the framework on synthetic data and real‑world applications, including intraday financial market volatility prediction and energy demand forecasting.
By Sophie Hanna Langbein, Niklas Koenen, Marvin N. Wright, Julia Herbinger
The paper investigates how different calendar feature encodings influence the performance of covariate-informed time series foundation models (TSFMs). Seven encoding schemes—three target‑agnostic (ordinal, one‑hot, cyclic) and four target‑based—are systematically compared across four state‑of‑the‑art TSFMs on nine datasets that reflect social activity patterns such as weekends and holidays. Results show that encoding choice can change forecast accuracy by up to 9.9 % on non‑holidays and 83 % on holidays, with improvements mainly around holidays and varying effects across models.
By Alexander Kreusel, Maximilian Beichter, Alexandra Nikoltchovska, Benjamin Sch\"afer, Ralf Mikut, Frederik Zahn, Veit Hagenmeyer
arXiv:2606. 27599v1 Announce Type: cross Abstract: While many explainable AI (XAI) methods have been proposed, most are not designed for time-series forecasting models and often rely on the implicit assumption that timestamp features are independent.
By Amadeo Tunyi
arXiv:2606. 01602v1 Announce Type: cross Abstract: Pairwise dependence measures such as correlation and causality are fundamental to temporal data mining, yet there is still no principled and robust way to quantify dependence between heterogeneous data types, especially between continuous time series and discrete temporal event sequences.
By Haoji Hu, Huaqing Mao, Yijun Lin, Xiaowei Jia, Jinwei Zhou, Minoh Jeong, Yao-Yi Chiang
arXiv:2608. 11508v1 Announce Type: new Abstract: Machine learning pipelines commonly flatten relational data into single-table representations, discarding structural constraints.
By Seungeun Lee, Joao Fonseca, Julia Stoyanovich